WEBVTT 00:00:00.000 --> 00:00:01.870 00:00:10.000 --> 00:00:10.590 00:00:13.320 --> 00:00:13.870 00:00:17.170 --> 00:00:25.960 Hello everybody, welcome to the January 18th ESE science similar and the. 00:00:26.630 --> 00:00:36.820 Announcements today are extremely brief. Next week, seminar will be given by Bel Air Lambert of UC Santa Cruz on absolute stress levels on fault. 00:00:38.020 --> 00:00:49.550 And then if you sign up for the Northern California Earthquake Hazards Workshop, just remember that's January 31st. But our registration is already closed, and those are all the announcements for today. 00:00:52.450 --> 00:00:52.920 So. 00:00:53.350 --> 00:00:57.100 And we're going to pass it to Jess Murray to introduce that, Chris. 00:00:58.250 --> 00:00:58.880 97. 00:00:59.250 --> 00:01:27.240 And so, yes, my pleasure to introduce Fred Pallets today, Fred, I think as as Evan has needs no introduction, but I'll give him one anyway. He received his Bachelors and masters from MIT and after that his PhD in geophysics from Princeton. And following that he completed postdoctoral work at several institutions both in the US and in Europe before coming to the USGS. 00:01:28.680 --> 00:01:57.360 His work has spanned a huge range of topics, just a few of which include induced seismicity, earthquake simulators, seismic velocity, structure of the crust and upper mantle. He's very well known for his extensive work on viscoelastic, modeling of post seismic processes and lower crust, upper mantle Realogy, and also inter seismic crystal strain rates and accounting for the long term transient effects. 00:01:57.450 --> 00:02:27.760 Umm that are seen in those in recent years. He was the project chief for the crustal deformation project here at the Earthquake Science Center and most recently has led the development of the next generation of crustal deformation models for the 2023 update of the national Seismic Hazard map. Today, he's going to be talking to us about several recent studies that he's carried out, in which he's developed, can amatic earthquake slip models derived from seismic in geodetic data. 00:02:27.870 --> 00:02:30.290 And it's my great pleasure to. 00:02:31.240 --> 00:02:32.780 I have friend here talking today. 00:02:35.970 --> 00:02:36.740 Thanks, Jessica. 00:02:37.590 --> 00:03:08.110 And thanks to the organizers for giving me the opportunity to give the seminar today, and it's just mentioned, this is based on a few published studies and my collaborators were Chuck Wicks and Jerry Schwartz from the Earthquake Science Center, as well as Bill Hammond, Thorne Lane and channel U. So I'm gonna be talking about four large intermountain W earthquakes that occurred in 2020 and 2021. And these are shown here. 00:03:08.620 --> 00:03:15.000 Together with a background map of the seismicity published by Steve Wisniewski. So. 00:03:15.590 --> 00:03:46.340 Of the first earthquake in this whole series happened on March 18, 2020, and that was right at the beginning of the pandemic and closely on the heels of that were the Stanley Idaho earthquake and the Monte Cristo Range, Nevada earthquake. And so that and the Antelope Valley earthquake in 2021 kept a number of us from different science centers very busy during the pandemic. And I'll just note for now that three out of these four. 00:03:46.660 --> 00:03:59.370 Earthquakes happened on faults that were previously not recognized, and three out of the four did not rupture the surface. Only the Monte Cristo range shows like rupture the surface. 00:04:04.100 --> 00:04:06.090 So. So what? 00:04:00.600 --> 00:04:06.770 So all of these earthquakes happened in places where size hazards were already estimated to be high. 00:04:06.530 --> 00:04:06.840 Good. 00:04:08.600 --> 00:04:13.150 Ohh, some of it's working but not the e-mail right now. 00:04:19.480 --> 00:04:20.170 Ohh sorry. 00:04:21.350 --> 00:04:51.760 So these earthquakes happen in areas where seismic hazard was already recognized to be high and a more detailed study of the hazard was given for the Wasatch Front area. You know, US fact sheet published in 2016 and there it was estimated that the 50 year probability of a magnitude 6.75 or greater was 43 percent 6.0 or greater 57% and then two 5.0 or greater with 93% so. 00:04:51.890 --> 00:05:03.130 Yeah, in that sense, the magnet earthquake is considered to be an expected event, but really all of these events are are no surprise. There's also historical seismicity in the vicinity of all of these events. 00:05:03.790 --> 00:05:13.450 So I'll I'll first dive into the magnet Utah earthquake 2020. This is a picture of downtown Magna before the earthquake. 00:05:15.310 --> 00:05:36.800 This earthquake had up to MI 8 damage and it did cause damage to a number of structures in Salt Lake City as well as magma and surrounding areas, so this is a picture of a fallen bricks on Main Street and Magna, and this is structural damage to a building in the western part of Salt Lake City. 00:05:38.790 --> 00:06:08.510 Uh, Steve was Nowski published a nice background figure for this earthquake, summarizing some of the recent seismicity, as well as the geodetic deformation and the fall team patterns. So the 2020 magnet earthquake was normal fall team events and the down dip direction of the Wasatch fall. The focal mechanism indicates East West normal faulting on a roughly 30 degree W dipping faults and the GPS. 00:06:08.720 --> 00:06:15.780 That's vectors where should kind of shown in blue. In this figure, they go from around 0 unstable North America to two to three millimeters per year West. 00:06:17.100 --> 00:06:37.530 West of the Wasatch Front. So they didn't. It's 2 to 3 millimeters per year extension across the range front. And I would add that another earthquake which isn't on this map was a 1962 event of roughly magnitude 5.2. That happened pretty much in exactly the same place as the Magna Earthquakes. So this area has. 00:06:38.770 --> 00:06:39.880 That earthquakes before. 00:06:41.240 --> 00:07:10.190 This is a a figure from a recent study by Emily Klaiber and others documenting the seismicity associated with the Magna earthquake. So the early aftershocks are shown in this large cluster to the West, which they believe is controlled by a solitaire grabbing structure that they document in their paper and a set of later aftershocks that started six or seven days later was located further east by the West Valley Fault Zone. 00:07:11.600 --> 00:07:19.970 And this is just another view of that seismicity from the same study showing the the Western cluster, which is also deeper. 00:07:20.390 --> 00:07:21.080 Uh. 00:07:21.910 --> 00:07:28.980 Uh, and more vigorous due to its proximity to the main shock as well as the Eastern Cluster located up to. 00:07:31.090 --> 00:07:45.000 So some of the questions that we have are what are the causative rupture plans, is the slip geometry consistent with normal faulty and observed geologically for the Wasatch Fault? Is there any geodetically? 00:07:45.630 --> 00:07:46.110 Uh. 00:07:47.270 --> 00:07:51.720 I measured after slip and how does it relate to the coseismic slip in depth? 00:07:54.490 --> 00:08:22.860 This is a a figure provided to me by Bob Smith and it's summarizes some of the patterns for previous large normal faulting events and I wanna highlight the plot on the lower right where it's it seemed that very shallowly and deep events very shallowly dipping and deep events are quite rare in the Magna earthquake does fit into that category. 00:08:24.380 --> 00:08:37.710 So the data set that I'm gonna use for this first study or seismic wave forms from the University of Utah seismograph stations using 3 component waveforms going out to 100 kilometer distance. 00:08:38.510 --> 00:09:07.520 Uh. Bandpass filtering pretty much up to a corner period of about four seconds using the 1st 36 seconds of all the records, and that's long enough to observe the principal signals prostatic offsets. The there is GPS. There are three sides with a resolvable signal, and there's also Sentinel 1, ascending and descending interferograms. So this is an example of some of the seismic data in 100 kilometer radius. 00:09:07.940 --> 00:09:15.110 From the events and we can see directly and especially direct access. It's really the direct S that's. 00:09:15.190 --> 00:09:24.350 Uh, showing up the most of the signal events is pretty much too deep to excite substantial surface waves close in. 00:09:25.770 --> 00:09:27.850 So this is for uh, different components. 00:09:31.510 --> 00:09:33.630 The geodetic data we have. 00:09:34.910 --> 00:09:58.390 Time series consisting of daily solutions for three different sites and the errors are are pretty high, but we can determine some offsets, at least for two of them that are well out of the noise. And as we'll see later, these offsets are consistent with normal faulting on a roughly NS trending fault. 00:10:00.460 --> 00:10:16.300 For interferograms, Chuck Wicks provided 2 interferograms for the study and the wrapped interferograms both indicate about 20 millimeters of subsidence and there are no sharp features on these interferograms which indicate the rupture is. 00:10:16.440 --> 00:10:18.410 Uh, pretty well buried. 00:10:19.770 --> 00:10:51.010 So for the slip inversion, I'm using seismic velocity structure developed defined by the Wasatch Fault community velocity model, and for this and the other studies that I'm gonna go into using simulated annealing for joint coseismic slip and after slip for this earthquake, I'm fixing the rake at the preferred solution determined by the University of Utah. There are positivity constraints on the on the slip and after slip, and there's smoothing applied to slip. 00:10:51.100 --> 00:10:59.500 Contributions and there's a specified apriori range of local rupture, velocity and maximum rise time. 00:11:01.070 --> 00:11:07.680 So their model issues that affect the study of the Magna earthquake as well as the other earthquakes that we're studying. 00:11:09.670 --> 00:11:27.660 Now the first is to decide on the fault geometry for coseismic slip and after slip I I find that in invariably there's a lot of of guesswork and just kind of manually adjusting what you think is the best geometry for. 00:11:29.020 --> 00:11:31.260 Sitting all the data that's available. 00:11:32.040 --> 00:11:34.850 Uh, sounds kind of old school to say it that way. 00:11:35.730 --> 00:11:58.830 That's and there are tradeoffs with choices of hypocenter limits on rupture, velocity, rise time, etcetera, and then the method is is limited by the choice of a single reference seismic structure, and so they're always path effects that are difficult to get a handle on, and I generally allow adjustments in the average. 00:11:58.910 --> 00:12:03.470 Ohh a path assessment of velocity to get around that. 00:12:05.250 --> 00:12:08.990 The vault geometry and hypocenter location for the magnet earthquake are provided. 00:12:09.610 --> 00:12:17.860 Uh, I study a pang and others in in 2020, basically indicating a. 00:12:19.970 --> 00:12:24.620 Ohh a hypocenter about 9 kilometers deep and. 00:12:25.610 --> 00:12:34.440 A dip of 34 degrees and a combination of of normal fault team and a right lateral slip component. 00:12:36.960 --> 00:12:55.030 The uh observed seismograms can be fit well with a model if they're on Hard Rock side, so this is an example of seismograms 3 component seismograms from a number of Hard Rock sites and the model that we come up with is in red and this fits the data. 00:12:55.550 --> 00:13:03.140 Uh, as as well as we can do with any site. However, there are a number of sites that. 00:13:05.300 --> 00:13:27.130 Are on areas of low seismic velocity very low seismic velocity as shown in the figure on the left, and that's results in seismograms that cannot be modeled. They're very strong base and effects and these and a number of other seismic Rams in these kind of red and orange regions are actually not used in the model. 00:13:29.000 --> 00:13:33.790 So we've specified the fault geometry and. 00:13:35.630 --> 00:13:53.550 I I should mention there there's a W dipping plane as indicated in the figure on the lower right and a steeply E dipping plane that intersects it at the hypocenter and we're resolving normal slip on either of these planes. 00:13:54.820 --> 00:13:56.150 Uh, we're also. 00:13:58.410 --> 00:14:26.360 Resolving after slip, which we suppose might occur on the the same plane which accommodated the coseismic rupture or may occur on the steeply NE dipping plan. And I'm finding that we can fit the data fairly well with a coseismic slip distribution as shown in the upper right. So the slipper originated at the hypocenter shown in the Yellow star and propagated. 00:14:26.440 --> 00:14:53.350 The to the North and South, but predominantly to the South, this was followed by after slip on the same W Emmy plane, but up dip of it. So when we plot the first few days of aftershock seismicity, we find that it it locates well to the east of the cost aspect slip, but it plots pretty much right on top of the after slip on the Westing plane. 00:14:59.680 --> 00:15:00.450 Hang it all. 00:15:02.080 --> 00:15:06.590 Address the question of whether Coulomb stress changes could. 00:15:08.520 --> 00:15:33.820 Explain that aftershock, seismicity and they immediate vicinity of the coseismic slip, as well as the later aftershock activity that was located further east and occurred several days later, and you can make up your mind looking at this figure, but quite a few of the events in both clusters do lie in areas of pause that have full on failure stress change. 00:15:35.900 --> 00:15:45.210 So how well does the model that I just presented the data so it fits the data pretty well at the Hard Rock sides as I indicated? 00:15:45.920 --> 00:16:15.490 Earlier, how well does it fit the geodetic data? So here I'm plotting the GPS vectors again together with the model projection based on just the coseismic slip or the combination of coseismic slip and after slip. And if we compare the upper left and the lower left plots, we see that we do need a component of after slip to explain at least the the two largest GPS vectors and. 00:16:15.990 --> 00:16:30.590 Yeah, the other GPS vector off to the east where we're not fitting particularly well, it has a very large error. And if we were absolutely determined to fit that vector, we would require more after slip. 00:16:30.950 --> 00:16:33.440 Uh, uh. Further up dip. 00:16:35.080 --> 00:16:40.790 About the same model does explain the interferograms pretty well with a small residual. 00:16:42.590 --> 00:16:46.160 There was an independent study by Masamori and others. 00:16:46.810 --> 00:17:10.180 Which use away field back propagation and when they use all of the available stations that includes the Hard Rock sites and other sites, then they find a Utah a unilateral rupture updip in towards the southeast. However, when they use a smaller number of stations which are basically just Hard Rock sites, they find a rupture. 00:17:11.560 --> 00:17:19.890 That that they describe this up dip into the South and that resembles a a little bit more my model which qualitatively. 00:17:19.980 --> 00:17:41.280 It resembles more unilateral rupture to the South, propagating more strongly to the South, to the and to the north. At any rate, the their cosmic rupture spills over to the east in the aftershock zone more than my model, but the southward rupture component is a common feature of the two models. 00:17:43.100 --> 00:18:12.430 So to conclude, for the Magna earthquake, the model is most consistent with the seismic waveform and the the static offset data involved. Coseismic slip on a shallowly W dipping false 9 to 11 kilometers depth, plus up to 10 centimeters of after slip on the same W dipping fault up tip of the locus of coseismic slip, and the event locates approximately on the down dip extension of the Wasatch Fault. 00:18:12.990 --> 00:18:34.500 And the shallow after slip reaching about 5 kilometer depth was likely triggered by the coseismic slip on the teeth or section of the Wasatch Fault, and I think these results are important just because we have this picture of cosmic slip happening deeper and after slip creeping up to shallow or levels. And the fact that the Salt Lake City segment. 00:18:35.660 --> 00:18:36.150 Is. 00:18:36.670 --> 00:18:42.300 Uh, it's it's very late, has not had a a rupture in a very long time. 00:18:44.600 --> 00:18:58.260 So next time I'm gonna change gears and move on to the March 31, 2020 Stanley Auto Horror earthquake. This happened just 13 days after the Magna earthquake and this is a picture of the Sawtooth Mountains, Idaho. 00:18:59.980 --> 00:19:02.790 I'll sleep with Nowski has another nice background figure. 00:19:03.480 --> 00:19:06.590 For this events and he, he notes that the. 00:19:07.710 --> 00:19:21.860 This event occurred in the western part of this at Centennial Tectonic belt, and in general activity in the Centennial Tectonic belt is concentrated between the Madison Range and the east, and the Sawtooth Range. 00:19:22.590 --> 00:19:23.410 In the West. 00:19:25.060 --> 00:19:33.880 There they are, 1959 Hebden Lake earthquake and the 1983 bore peak earthquake happened when the Centennial tectonic belts. 00:19:34.540 --> 00:19:40.150 Uh. Moving to the sawtooth, fall to slip, rate of .9 millimeters per year is estimated. 00:19:41.710 --> 00:20:09.770 And activity on the sawtooth false is considered to have migrated to the north during the Pleistocene. All the action during this earthquake was off the northern part of this false net extension perpendicular to the normal faults from GPS, is about when millimeter per year and the area undergoes right lateral share between .3 and 1.5 millimeters per year, and that would be right lateral shear across this. 00:20:10.570 --> 00:20:15.620 Uh NE trending Centennial tectonic valves and. 00:20:16.050 --> 00:20:17.480 US West nowski. 00:20:17.570 --> 00:20:28.880 He states that this is consistent with the the left lateral or the left the stepping and that and that salon pattern that we see in a number of these faults systems. 00:20:29.990 --> 00:20:33.890 So this is a rock slide in the Snake River Canyon. 00:20:34.770 --> 00:20:41.710 Due to this event it's it's not easy to find a lot of damage because the areas remote and it was pretty deeply buried. 00:20:43.490 --> 00:20:56.170 This is an example of a sandy beach which used to be here and is labeled former beach due to liquefaction features and this is an example of a tree which fell over again due to liquefaction. 00:20:57.450 --> 00:21:01.020 So I'm showing here again the Centennial. 00:21:02.060 --> 00:21:32.730 Tectonic belt and uh Stanley earthquake occurred off on the West western part of that. So the questions are what are the causative rupture planes? Was there surface rupture? Is the slip geometry consistent with normal faulting observed geologically for the Sawtooth fault and how does the trends child's fault system or influence earthquakes slip? So it has been noted before that the trends Challis fault system tends to terminate all these N Northwest. 00:21:32.880 --> 00:21:33.560 Trending. 00:21:35.100 --> 00:21:42.920 Oh oh, more recently active fault systems. The trans trella Trans Challis fault system. 00:21:44.240 --> 00:21:51.180 What dates back to the Essene and it's a it accommodated normal faulting at that time, so it's an ancient fault system. 00:21:54.110 --> 00:22:22.760 But these figures by liberty and others 2020 indicate that the the seismo tectonics are dominated by a combination of normal faulting and left lateral slip, and they also note looking at the aftershock seismicity that the likely cause the default plane was a steeply W dipping structure, and they also considered an interferogram which indicates subsidence. 00:22:23.200 --> 00:22:31.190 To the West and uplift to the east, which is consistent with a combination of normal slip and left lateral strike slip. 00:22:32.250 --> 00:22:42.340 So to further investigate this earthquake, I'm using a data set of seismic wave forms from several networks going out to 300 kilometers. 00:22:42.910 --> 00:22:44.370 Uh away. 00:22:44.450 --> 00:22:52.240 Uh for this event and using somewhat lower frequency, so down to a corner period of about 8 seconds. 00:22:53.020 --> 00:23:08.470 Using 120 seconds of all the records and for static offsets we have horizontal offsets from regional continuous GPS sites as well as in ascending and descending Sentinel one image. 00:23:13.040 --> 00:23:23.750 So on the seismic data that I mentioned is within a 300 kilometer radius is is shown here, there are 18 regional stations. 00:23:25.020 --> 00:23:34.950 And I mentioned these other things already, so the coverage is is not great in the immediate vicinity of the event, which is why we need to go to. 00:23:36.110 --> 00:23:37.520 Longer period and. 00:23:38.340 --> 00:23:41.600 Extend the range further out compared with the magnet earthquake. 00:23:42.830 --> 00:23:48.630 Are GPS data. This was provided by the Nevada Dudik Laboratory. 00:23:49.570 --> 00:24:01.740 And this pattern is roughly consisting with a combination of of normal fault again and left lateral strike slip faulting. And when you just look at this pattern. 00:24:02.220 --> 00:24:11.550 So it's it's more consistent with a strike slip faulting than than normal faulting, even though both are are taking place. 00:24:12.580 --> 00:24:16.880 During this event and we can fit this data with a pretty healthy combination of both. 00:24:18.370 --> 00:24:32.200 So this is just showing that the the GPS offsets are based on data collected several days before and after the event and we can see pronounced steps for some of these stations. Some of these larger offsets. 00:24:33.290 --> 00:24:36.800 Close and we see the steps clearly in the data. 00:24:39.630 --> 00:25:07.180 So we've got uh Sentinel 1 ascending and descending interferograms and it it boils down to about 50 millimeters of subsidence to the West and an equal amount of of uplift to the east. But it it is shaped by a combination of normal faults, pain and left lateral slip. And the same goes for the dissenting interfere. 00:25:10.140 --> 00:25:17.430 To get a handle on the fault geometry, we can make use of either the any IC preferred moment sensor or the. 00:25:21.010 --> 00:25:34.340 C GMT solution which are are fairly close to each other and define and nearly NS trending fall plane which I have dipping steeply to the West based on the. 00:25:34.480 --> 00:25:37.080 Uh, aftershock pattern. 00:25:38.770 --> 00:26:01.020 I've plotted by liberty and others, and we're allowing for a combination of left lateral strike slip and normal slip in this case. So variable rake and we're also allowing for a combination of possible right lateral strike slip either cosmetically or cosmically or in the form of after slip along a couple of. 00:26:01.670 --> 00:26:22.820 Additional planes that I've labeled fault two and fault three. That's to recognize their lineations in the aftershock pattern that seems to line up with the trans child's fault system. So it's suggestive that maybe something was activated along the ancient fault system, but out of curiosity. 00:26:23.920 --> 00:26:28.710 We would like to have an idea whether Coseismic slipper after slip was accommodated. 00:26:30.550 --> 00:26:36.240 So the slipping version is similar to the procedure for the Magna earthquake. 00:26:37.360 --> 00:26:43.830 The only difference is we're allowing for variable rate in this case, rather than fixing the rake. 00:26:46.130 --> 00:26:47.300 These are the results. 00:26:49.030 --> 00:27:05.580 That we're getting for the net slip, so this is a steeply dipping plane so that we see on the left of this is the coseismic slip and the rupture begins with a predominantly left lateral slip, but it transitions. 00:27:06.330 --> 00:27:10.390 To predominantly normal slip in the South, which is consistent. 00:27:11.190 --> 00:27:30.900 With overlapping the the northern portion of the Sawtooth Fault, which is geologically recognized as a normal fault, we allow for coseismic slip on the these other two fault planes and I'm just showing the cosmic slip inferred on this. 00:27:34.410 --> 00:27:42.780 Fault #2 playing as an example and there's very little of it. However, if we allow for after slip. 00:27:43.590 --> 00:27:52.200 On those two planes, we find that the F2 plane does accommodate a significant amount of after slip around 20. 00:27:52.950 --> 00:27:57.000 Ohh centimeters. But it is a depth. I should mention these. 00:27:57.790 --> 00:28:11.500 Uh planes that are define go from the surface to 20 kilometers depth, so looking for example at the F2 plane this after slip would be 10 kilometers or or deeper. 00:28:15.720 --> 00:28:37.020 When we look at the cosmic rupture, how how this this slip progressed with time, we see that it did indeed start with predominantly left lateral slip in the north, but it it's migrated quickly and really gained steam between 6:00 and eight seconds and transitions to normal slip and this is. 00:28:38.430 --> 00:28:40.600 Just another view of the. 00:28:42.260 --> 00:28:45.680 Net coseismic slip distribution. So we have. 00:28:55.880 --> 00:29:05.880 And this is how well we're fitting the seismic data. And I always think it's it's not perfect and how could we do better, but it's doing a decent job of this event. 00:29:07.510 --> 00:29:13.210 Similarly, we're explaining the GPS data as as well as well as we can. 00:29:14.910 --> 00:29:22.660 There are fits are not sensitive to the details of the faulting. Obviously there are only seeing the big picture and we're fitting them reasonably well. 00:29:23.850 --> 00:29:35.220 There is an alternate model for this earthquake and also just jump to the next figure. It involves a coseismic slip not only on this. 00:29:36.010 --> 00:29:40.760 Red Plain, which is very similar to the plain that I defined. 00:29:41.720 --> 00:29:57.900 For microseismic slip model, but they invoke another plane off to the West, which dips in the opposite direction. So we have a A W dipping plane to the north shown by these red segments, and in East dipping plane. 00:30:00.640 --> 00:30:12.500 Off to the South and the West, which overlaps the West Dipping fault at shallow depth. However, the the slip distribution. 00:30:23.460 --> 00:30:43.560 Ohh are trying to decide if this model is plausible or not, but this model does have some attractive features. It does account better for some of the aftershock seismicity that's located off to the South and it does fit the interferograms better than the model that I presented earlier. 00:30:46.280 --> 00:30:55.950 So to conclude, for the Stanley earthquake, the event was a unilateral rupture with a bleak slip beginning with left lateral strike slip in the north and transitioning to normal slip in the South. 00:30:56.570 --> 00:31:05.580 The rupture may have jumped up to 10 kilometers from a primary fault dip into the West and a second fault dip into the east and parallel to the Sawtooth Fault. 00:31:06.590 --> 00:31:12.920 And after slipping, might have occurred along the east and NE trending chalice fault system. 00:31:16.990 --> 00:31:24.080 I'll move on to the money Crystal Range, Nevada earthquake. This is the only surface rupturing advance out of the four. 00:31:27.000 --> 00:31:45.480 And Steve Lesneski has a another nice background figure for this event and he notes that it occurred north of the 1872 Owens Valley earthquake as well as South of several notable earthquakes in the Central Nevada Seismic Zone. 00:31:49.830 --> 00:32:03.370 And it occurred in a right step within the northern Walker Lane, and this step is referred to as the Mina deep deflections, which accommodates regional transcension it. 00:32:04.380 --> 00:32:11.590 The event basically ruptured 3, allowing E trending false that overlap the Candelaria fault and continue to its E. 00:32:13.990 --> 00:32:18.940 OK. Ohh color risk color and others and Steve Wazowski. 00:32:20.180 --> 00:32:36.340 Beforehand. Note that the transfer of right lateral slip in the mean of deflection is accommodated by clockwise rotation. A crustal blocks bound by east striking left lateral faults. So Wasniewski proposed this in 2005 and Kohler and others make. 00:32:37.530 --> 00:32:45.360 Several strong geologic arguments for why this is correct, so the causative faults in the the East West trending faults. 00:32:47.070 --> 00:32:56.840 Uh, that ruptured in the Monte Cristo range. Earthquake would correspond to these small left lateral ruptures in these rotating blocks. And then this figure. 00:32:59.780 --> 00:33:02.550 This is the GPS data available. 00:33:03.380 --> 00:33:17.020 For this earthquake, it's consistent with predominantly left lateral slip on an E Westing East West trending falls, there are two interferograms that we're using in this study. 00:33:19.990 --> 00:33:22.100 I would note that the the. 00:33:23.610 --> 00:33:30.960 Strong blue region that you see in the ascending interferograms the product of subsidence as well as. 00:33:31.560 --> 00:33:42.120 Ohh eastward motion due to the left lateral slip and in the ascending image the this the same blue region is due to a a very strong down drop. 00:33:44.080 --> 00:33:46.470 Whereas the red region is due. 00:33:47.400 --> 00:33:59.170 To the effect of eastward motion or the left lateral slip component. So this down drop basically implies that there's a minor amount of normal faulting in in addition to the left lateral slip. 00:34:00.170 --> 00:34:01.520 Which dominates the earthquake. 00:34:16.010 --> 00:34:18.250 So to construct the kinematic slip model. 00:34:19.530 --> 00:34:50.830 Uh, Chengli Liu and colleagues decided on three fall planes with upper lower edges, strikes and and dip constrained by aftershock, seismicity and insar, so the the aftershock seismicity is pretty broad, and you could probably draw lines almost anywhere you want. When I look at it like this, but it is a constraint so that the the two vertical strikes that falls would be shown by these. 00:35:27.290 --> 00:35:40.760 So this is the slip distribution that we obtain where when we consider the the joint seismic waveform, GPS and INSAR data, the slip. 00:35:41.440 --> 00:35:58.600 Is concentrated mostly below 3 kilometers depth, but some of the slip does reach close to the surface directly above the high slip region and I'll say more about that in a moment. Most of the the slip. 00:35:59.280 --> 00:36:06.100 Happened between 5 to 10 seconds into the rupture. The the rupture migrated from. 00:36:07.920 --> 00:36:15.070 East to West predominantly as a unilateral rupture, although there is some coseismic slip off to the east as well. 00:36:17.340 --> 00:36:23.120 So here I'm trying to connect the slip at depth with what was observed. 00:36:23.890 --> 00:36:33.260 At the surface, so the detailed study by Collar and others measured fractures that that occurred at the surface. 00:36:38.910 --> 00:36:39.500 No. 00:36:41.820 --> 00:36:50.500 Left lateral E trending fault that accommodates a lot of slip. Rather there are fractures the accommodating. 00:36:51.310 --> 00:36:54.500 A left lateral slip trending roughly north eastward. 00:36:55.390 --> 00:37:06.770 Uh in the western part of the rupture, and there are other fractures trending NS, accommodating right lateral slip off in the eastern part of the rupture and I've. 00:37:07.540 --> 00:37:10.720 Indicated by eye where the. 00:37:12.230 --> 00:37:13.540 The surface map. 00:37:14.520 --> 00:37:15.190 Mapping. 00:37:17.170 --> 00:37:18.820 From the field geology. 00:37:24.980 --> 00:37:50.910 These fractures were seen at Earth. Surface coincides well with the area of the greatest amount of slip that we see at depth, and arguably even that the highest density of fractures, if you will, is located above the area where we have the largest coseismic. Slip it depth and I think this is seen in a lot of earthquakes. We saw a similar thing in the Ridgecrest earthquake. The maximum surface slip was right about. 00:37:58.900 --> 00:38:04.790 There have been a few studies of the after slip from this event and this. 00:38:07.740 --> 00:38:29.550 Uh, they solve for both the coseismic slip and the after slip, and they found, as did we, that the coseismic slip is concentrated between 3 and 12 kilometers depth and the after slip, which is given by the colored pattern in these plots located primarily shallower than the coseismic slip. And that's true both in. 00:38:30.500 --> 00:38:35.550 The the kind of the eastern domain and the western domain. 00:38:36.260 --> 00:38:42.650 But they they are finding quite a bit of after slip near the surface, so I would. 00:38:43.320 --> 00:38:46.680 Put the question to the field geologist if they've seen. 00:38:47.420 --> 00:38:55.340 That, uh, that amount of after slip, it's our any indicators that there's that that amount of after slip at the surface. 00:39:13.130 --> 00:39:19.620 The the slip is distributed between 3 and 15 kilometers depth and the largest shallow slip. 00:39:20.490 --> 00:39:27.860 Near the transition between the left lateral strike slip to the east and the normal slip to the West is coincident with most of the observed surface slip. 00:39:28.770 --> 00:39:36.680 And the earthquake occurred when the within his own of rotating blocks bounded by the east West strike, slip faults in the mean of deflection. 00:39:39.790 --> 00:39:46.800 So I'll move on to the 2021 Antelope Valley, California earthquake. This is a picture of the Antelope Mountains. 00:39:49.210 --> 00:40:08.960 So this earthquake was a normal faulting event in the Central Walker Lane, which accommodates 20 to 25% of the dextral shear between the Pacific and North American plates, and the epicenter locates below the Antelope Valley Fault zone, which has an estimated length of 23 to 30 kilometers. So that. 00:40:09.630 --> 00:40:16.740 That fault zone presents its own substantial seismic hazard, although that particular fault did not rupture in this event. 00:40:42.370 --> 00:41:00.660 Ohh, the the rupture propagated unilaterally along strike and up dip, and it our interpretation the fault plane is located somewhat below the locus of seismicity and that's. 00:41:01.180 --> 00:41:14.500 Ohh, kind of a bone of contention. Whether that generally happens or whether the fault plan should be drawn right through the the middle of of the cloud of of aftershocks. But I'll return to that question. 00:41:15.420 --> 00:41:28.420 So some of the questions are just one plane self host to capture the coseismic slip is the cosmic slip distribution uniquely constrained by the seismic waveform and geodetic data? Did the earthquake nucleate rapidly or slowly? 00:41:29.540 --> 00:41:46.770 So again, we're using the data set of seismic wave forms and side static offsets from GPS and INSAR and for this event using data out to 100 kilometers and pushing it to about 6 seconds. 00:41:50.840 --> 00:41:53.470 So these are the contributing seismic wave forms? 00:41:54.090 --> 00:42:04.950 Ohh, there's a large concentration of contributing stations to the North, but there's also cluster of stations near Long Valley which do contribute valuable data. 00:42:06.990 --> 00:42:08.270 This is an example. 00:42:09.010 --> 00:42:12.310 Of a GPS offset at a. 00:42:13.250 --> 00:42:29.690 A A continuous site LNT which by serendipity lies right above the rupture, so it has very large horizontal and north vertical steps that you can clearly see in the daily time series. 00:42:32.530 --> 00:42:42.090 There are a number of interferograms available for this event, and they they indicate roughly 60 millimeters of subsidence. 00:42:42.750 --> 00:42:44.050 Associated with the events. 00:42:46.880 --> 00:42:53.390 Well, this or these. These two scenes are are clearly noisier, but they still show. 00:42:54.230 --> 00:42:59.690 The same pattern of subsidence and these descending paths. 00:43:00.720 --> 00:43:04.430 Uh, similarly show in this case similar amounts of. 00:43:05.090 --> 00:43:06.410 A line of sight change. 00:43:08.290 --> 00:43:15.640 So in the slip inversion, we're using the seismic structure provided by Mancino and others 1993. 00:43:16.200 --> 00:43:19.960 Oh oh, show you. 00:43:22.240 --> 00:43:26.850 Uh, some evidence for a relocated hypocenter being located. 00:43:27.590 --> 00:43:39.930 Uh, quite some distance away from the preferred Nic hypocenter, and we're applying the same methodology as before. So similar to the magnet earthquake I'm using. 00:43:41.030 --> 00:43:42.600 I'm I'm specifying a constant. 00:43:43.430 --> 00:43:46.080 Paul geometry, including fixing the ring. 00:43:48.300 --> 00:43:50.110 So for for this event. 00:43:51.440 --> 00:44:02.470 In trial models, I found that, uh, the model for that I explored preferred nucleation in the North, and that was at odds with the Nic. 00:44:03.560 --> 00:44:11.290 I preferred hypocenter, which is located further South of it. However, if one examines the peak travel times. 00:44:12.410 --> 00:44:16.960 They are more consistent with a hypocenter to the north. 00:44:17.340 --> 00:44:20.320 Uh. Uh then? Otherwise, so I. 00:44:21.230 --> 00:44:24.240 Decided to keep it. Where are the model seem to prefer it. 00:44:25.700 --> 00:44:45.270 So the fault geometry as I mentioned before, there's a 50 degree dipping fault plane which plots slightly below the the locus of size unicity, and the main reason for that is the model tends to overpredict the amplitude of the subsidence measured by insar and. 00:44:45.960 --> 00:44:49.440 If I specify a shallower plane. 00:44:50.150 --> 00:44:56.030 That's misfit, grows even larger. So there's a trade off there. 00:44:58.900 --> 00:45:08.960 And this is the model that we come up with and there there is some noise to these models. This is only a 6.0 and. 00:45:11.080 --> 00:45:17.670 There are no extremely nearby seismic stations, but basically if we look at the. 00:45:18.780 --> 00:45:50.730 Part of the ruptured time as a function of time. This indicates ***. We're progressing rupture, however, significant slip really didn't happen until about 3 seconds after the rupture. Tom in this model, so that indicates a slow nucleation process for this event, and that's implied by the figure on the right as well. But overall the the rupture propagated along strike and dip, but did not come anywhere near the surface in fact. 00:45:50.850 --> 00:45:56.540 We're cutting off our slip distribution candidate slip distributions at 4 1/2 kilometer temps. 00:45:58.380 --> 00:46:06.830 So this model fits the seismic waveform data pretty well. There are some exceptions where we have side effects that are clearly going on. 00:46:11.210 --> 00:46:16.120 I'm showing all of the seismic wave forms, including the Long Valley stations, so. 00:46:19.550 --> 00:46:29.750 So this is the fit to the GPS static offsets on the left and it's fitting fairly well including the very large offset that we looked at at LANT. 00:46:31.050 --> 00:46:46.830 The 5th of the line of sight displacements from inside is pretty good, but three out of the four interferograms some are summarized here the they're kind of over predicted by the model, and I never found a really satisfactory. 00:46:47.620 --> 00:46:52.030 Way out of that, there's probably a little something extra. 00:46:52.680 --> 00:46:57.490 Going on that I did not quite get a handle on. However, there is an independent. 00:46:58.890 --> 00:47:08.630 Model for this earthquake, which I'll get to maybe in the next slide. Yes, this is it. So Tang Wang and Doug Dreger have an independent model. 00:47:09.550 --> 00:47:22.480 Uh and their candidate plane is similar to mine, but it looked it locates through the middle of the aftershock seismicity, and they do manage to fit the interferograms. 00:47:23.370 --> 00:47:26.320 Ohh, with this model I'm not quite sure. 00:47:27.180 --> 00:47:30.280 How they did it, but it's always good to have. 00:47:31.180 --> 00:47:44.850 Alternate models and I'm I'm eager for this study to be published so that I can learn all the details. What what did they do that I couldn't do? But it is. It is an alternate model. 00:47:45.190 --> 00:47:50.280 Ohh, qualitatively they're getting a similar slip distribution. 00:47:51.400 --> 00:48:10.830 To mine. Basically the rupture propagated along strike and updip. However, it the rupture begins at the end I see hypocenter where his mind begins further N to the substantial coseismic slip, you might say, starts maybe a couple of kilometers to the northeast of theirs. 00:48:11.530 --> 00:48:39.190 There are big trade-offs in this kind of work with the assumed seismic velocity structure, and that by itself can explain these kinds of of differences as well as the the hypocenter location that's chosen as well as the fact that I'm using velocity waveforms in this kind of modeling, whereas most other practitioners, including Wang and Dreger use displacement waveforms. 00:48:41.620 --> 00:48:57.150 Nevertheless, the the segment moments are are pretty similar and the overall rupture pattern is similar. There's just kind of displaced a little bit with respect to one another. They they both agree the the slip is is pretty deeply buried. 00:48:58.140 --> 00:49:10.490 So to conclude, for the Antelope Valley earthquake, it was a unilateral rupture with predominantly normal slip on a he slipping fault propagating to the South and up dip and the slip is concentrated between 7 and 10 kilometers depth. 00:49:12.160 --> 00:49:14.320 The aftershocks which I I showed. 00:49:16.130 --> 00:49:41.180 Many, many slides ago, they extend up dip to about 1 kilometer depth, which is much shallower than the coseismic slip. So possibly there is shallow after slip, but nobody that I know of has investigated that and I'm finding there's about a 3 second delay between the origin time and the onset of significant main shot slip which was suggested an emergent event with an initially slow nucleation process. 00:49:43.610 --> 00:50:12.830 So to summarize, for the the four Intermountain W Earthquakes 3 out of the four occurred on previously unknown faults. Only. Arguably, the magnet earthquake occurred on the down dip extension of the Wasatch Fault. All four occurred in areas of relatively high background seismicity, and geodetic strain rates, and these events sample different tectonic environments with distinct faulty geometries, leading to you unique ruptured. 00:50:12.900 --> 00:50:14.350 Characteristics of each event. 00:50:18.180 --> 00:50:19.080 Thanks for your attention. 00:50:27.930 --> 00:50:30.320 Any questions in the room before we go to the chat? 00:50:32.150 --> 00:50:34.190 Yeah. So earlier you mentioned. 00:50:35.030 --> 00:50:47.470 But actually I guess maybe it's from Magna and remember exactly, but I think it was, Stanley, Stanley, OK, do you see evidence of that on the interference and does that potentially affect your after sleep? 00:50:48.650 --> 00:50:52.050 Restaurants or, yeah, Chuck Wicks looked at the. 00:50:53.120 --> 00:50:55.470 Interferograms for the magnet earthquake. 00:50:56.790 --> 00:51:00.020 If we're trying to the Mega earthquake and. 00:51:01.700 --> 00:51:07.160 There are liquefaction effects and I guess lateral sliding in the Kennecott. 00:51:08.250 --> 00:51:24.350 Tailing minds pawns off to the West of Magna, and that was masked out in the interferograms. OK that we use, but yeah, otherwise I wasn't really cognizant of liquefaction features and the other interferograms that we analyze. 00:51:27.420 --> 00:51:34.180 So Fred, that was a great talk and now I have a question. So a number of your, YOUR solutions showed shallow, obviously. 00:51:34.830 --> 00:51:57.590 And you think that's a consequence? Are these range bounding faults? Are there Robbins containing soft sediments for example? In other words, I was looking is there something analogous to what we saw in Napa earthquake where you get the after slip and the clay Ridge soft sediments and the coseismic predominant mania hard, harder velocity, you know weakening materials. Do you see anything like that happening here or have you thought about that problem? 00:51:59.910 --> 00:52:08.120 You know, phenomenologically, this seems to happen over and over. The coseismic slip is fairly deep. And then the after slip is is shallower, so. 00:52:09.850 --> 00:52:26.130 I think I touched on that for the magnet earthquake, there's they have to slip with shallow or they'll go nowhere near earth surface. And for the Monte Cristo Ranger of quickly after slip seemed to fill in, fill in the slip deficit at shallower depth and depth. 00:52:27.090 --> 00:52:39.690 So. So what I was wondering is there is there independent evidence that there might be sediments up against one side of the ball, for example, in the hanging wall for a normal event from you know people have a pretty good idea of the structure of some of these bases. 00:52:40.540 --> 00:52:46.550 I'm thinking is there any independent evidence for sediments extending to 2-3 kilometers that would be pretty deep but not unheard of. 00:52:47.640 --> 00:52:52.170 Certainly, yeah. For the Magna earthquake case, if we. 00:52:53.290 --> 00:53:05.460 Yeah, we suppose after slip code have extended as far up dip as the West Valley Fault zone, but that would that would be still depths of say 5 kilometers which is below. 00:53:06.160 --> 00:53:24.610 The sediment depth even in the the different basins over there. So yeah, like that's why 3 kilometers might be limit. Yeah, it doesn't quite apply. OK, yeah. So something else is going on. Something to keep in mind. OK. Just there's race strengthening material there. That's what it is. It doesn't have to be sediment. So that's good. 00:53:26.740 --> 00:53:41.330 Yeah, that was that was fun, friend for the Magna event. I think that's the first one you talked about, right? So you showed a bunch of signs or grams that you couldn't fit with high amplitudes. So I'm curious is that those seems like the side effects from the low velocity. 00:53:41.980 --> 00:53:59.230 Sentiments or something just is that your interpretation? Is there? Yeah. Yeah, it is that a limitation of the velocity model that you have in this area or and and then what like what percent of the stations were that was that that you weren't able to fit in with that change anything if you could use those? 00:54:00.720 --> 00:54:05.070 Yeah, those are good questions. Those stations are very much correlated. 00:54:05.690 --> 00:54:12.550 With the presence of like I, I was calling VS500 like five 500 meter deep. 00:54:14.170 --> 00:54:14.860 PS. 00:54:15.910 --> 00:54:16.530 Uh. 00:54:17.500 --> 00:54:21.780 And a previous study by Morgan Machete also identified that. 00:54:22.420 --> 00:54:29.680 Waveforms are more complicated above the the same basins. He was also working with the Wasatch Community velocity model. 00:54:31.550 --> 00:54:48.520 Whether or not to use those sites, I'm limited by assuming A1 dimensional seismic velocity structure and I I will allow for path dependent. Just velocity changes independent of depth, so it's a very crude correction for path dependent velocities. 00:54:49.680 --> 00:54:50.780 Station by station. 00:54:51.540 --> 00:54:52.320 But that. 00:54:53.070 --> 00:55:00.380 Would not handle these kind of side effects. It would be nice to include more of the stations. I think about 40% of them. 00:55:01.180 --> 00:55:01.880 Uh, yeah. 00:55:02.570 --> 00:55:05.680 Thrown out because because of side effects. 00:55:12.170 --> 00:55:16.180 Is there any questions online? I can't see if people are raising their hands. 00:55:16.080 --> 00:55:19.210 Yeah, we have a few questions in the chat. 00:55:19.790 --> 00:55:34.710 Umm Jessica Marie asked. You mentioned how these events mostly ruptured, unknown faults, but we're in areas of high background seismicity and strain rate. Could you comment on how this sort of deformation is accounted for in the recent NSHM update? 00:55:37.100 --> 00:55:42.890 Uh, So what does this mean for in the context of the recent NSHM update? 00:55:43.960 --> 00:55:59.000 Well, I'll leave assembled for the yeah, the tectonic geology group, which has been led by Alex Haddam for the Western US assembled, 1017 volts for for that update. 00:55:59.400 --> 00:56:13.720 Uh, so this would be an argument that any compilation like this is incomplete, though we keep seeing events after event that occurs on a previously unknown fault, but that that effort is is very much worthwhile because. 00:56:14.510 --> 00:56:27.540 The events are clerk per close to false that we know about, and so hazards are are defined in the general vicinity of faults, not just exactly what I'm on one fault so. 00:56:29.360 --> 00:56:35.010 But all of these events already occurred in areas where the the hazardous considered to be high based on. 00:56:35.980 --> 00:56:40.950 Pre previous seismic historical seismicity and current. 00:56:41.670 --> 00:56:42.200 Uh. 00:56:43.550 --> 00:56:50.450 Seismicity measured with the modern instrumental network, as well as the geodetic string. 00:56:56.720 --> 00:57:11.150 Also, we have another question from Belle Philibosian. Perhaps you said this and I missed it. Does the Animal Valley fault plane not seem to be contiguous with the previously mapped Linker Valley Fault which was at least initially thought to be the likely source of that event? 00:57:14.590 --> 00:57:23.950 I didn't mention the Slinker Valley fault. If that has been associated with the Antelope Valley earthquake and the. 00:57:24.900 --> 00:57:28.990 The proposed faults that either I or Tang Wang. 00:57:29.810 --> 00:57:31.790 Are employing would. 00:57:32.650 --> 00:57:41.290 Come out roughly at the surface trace of the Slinker Valley Fault, but it's actually hard to find information about this Slinker Valley fault, and it's not one of the faults in. 00:57:42.100 --> 00:57:44.440 This round of the NSA charm, as far as I'm aware. 00:57:50.640 --> 00:57:56.110 Thanks. Do we have any hands? I don't see any more questions in the chat. 00:57:57.270 --> 00:58:03.650 And also don't see any ohh Sue Huff. Just raise your hand so you wanna unmute yourself. 00:58:05.580 --> 00:58:08.640 Yeah, it's a bit like. Thanks for the talk, Fred. I'm just. 00:58:09.530 --> 00:58:14.080 Wondering is is this a normal rate of activity for the region? All these sixes? 00:58:24.070 --> 00:58:26.190 They NSHM work. 00:58:27.180 --> 00:58:34.730 Of the the deformation group that I LED, as well as the overall effort that Ned is leading the. 00:58:34.810 --> 00:58:42.580 The the moment the accumulation rate in the western US for a long fault, So what it is is 2 to 10. 00:58:43.620 --> 00:58:51.890 Between 2:00 and 2 1/2 * 10 to the 19 Newton meters per year. All fault deformation would kind of raise that to 3 * 10 to the 19. 00:58:54.030 --> 00:59:01.720 Newton meters per year, which I think boils down to roughly a magnitude 6.5 equivalent every every year. 00:59:03.560 --> 00:59:03.980 OK. 00:59:02.440 --> 00:59:09.870 So we had, yeah, roughly four of them in two years. So it's a little bit about normal, but but we have slowed down. 00:59:11.860 --> 00:59:16.060 OK, I realized it was outside the scope of your talk. It's just just wondering. Thank you. 00:59:23.720 --> 00:59:26.400 Anymore questions from the audience? 00:59:28.230 --> 00:59:51.780 Umm. If not, I think we're at 11:31, so maybe we can stop recording and we can leave the video on in case there are people online who wanna stay in chat. But Fred is actually in the user menu room. So if you're in Moffitt, I'd encourage you if you were interested in chatting later about the talk. So just go down. 00:59:53.390 --> 00:59:54.060 And then.